This work proposes Reality-Anchored Object Memory Architecture (RAOMA), a framework for structuring AI knowledge around persistent real-world objects anchored in space and time. Instead of treating knowledge as fragmented symbolic or statistical patterns, RAOMA organizes data as object-centered timelines with interaction histories and confidence-evaluated knowledge attribution. The architecture introduces spacetime indexing, causal event logging, and AI-mediated validation to improve reality correspondence, reduce redundant retraining, and support long-term knowledge preservation. A multi-layer governance model balances public knowledge sharing with sovereignty and privacy considerations. RAOMA is positioned as a potential civilization-scale knowledge infrastructure supporting AI reasoning, scientific discovery, and institutional memory
Moutsopoulos et al. (Fri,) studied this question.